Practice Profile: Skill-Building Interventions for Delinquent Behaviors of Youth

Tennyson (2009) conducted a meta-analysis to examine the impact of therapeutic treatments aimed at reducing juvenile recidivism. Databases (including PsycINFO, ProQuest Dissertations and Theses, Social Services Abstracts, Sociological Abstracts, and MEDLINE) were searched to identified eligible studies between 2003 and 2008. To be included in the meta-analysis, studies had to examine justice-involved youth ages 12 to 21; use an experimental or quasi-experimental design; be published in English; and examine at least one intervention aimed primarily at reducing juvenile recidivism (which was defined as rearrest, reconviction, or reincarceration of justice-involved youth). Overall, the interventions could focus on a variety of treatment modalities, such as individual, family, group, or multisystemic therapies and correctional programs, parent training, peer influences, or restitution. (The CrimeSolutions review of this meta-analysis focused only on the studies of skill building programs, n = 9). Treatment also could take place in an inpatient or outpatient setting and could use an assortment of therapeutic orientations such as cognitive–behavioral, behavioral, or integrative therapies. Youth could participate in the interventions during incarceration, while on probation, or during aftercare (following release from incarceration).

More than 1,000 studies were identified in the initial search; of those, 33 were determined to meet the eligibility criteria. An additional study was identified when searching the reference section of the eligible studies. Of the 34 studies, 4 would be removed because of issues with inadequate data provided (to calculate effect sizes), leaving a final sample size of 30 studies. The 30 studies included more than 100,000 juveniles. The average age of youth was 16 years (they ranged from 14 to 20 years). More than 50 percent of the sample of youth from the studies were White, approximately 30 percent were Black, and the remaining participants were identified as Hispanic, Asian, American Indian, or other ethnic origin including mixed race. More than 80 percent of the sample of youth were male.

Of the 30 studies, 73 percent were published articles and 27 percent were unpublished dissertations. About 33 percent of the studies randomly assigned youth to treatment groups, 63 percent used a nonrandom design, and in 4 percent of studies it was unclear the type of research design employed. Approximately 23 percent of the treatment groups in the included studies received care in a detention center; however, most of the treatment groups received treatment in other settings (67 percent), including in the community or at home. With regard to the treatment domain of the interventions, 13 percent of studies used group, 3 percent used family, 61 percent used multiple types (meaning a combination of therapies such as individual, group, or family), 13 percent used other types, and in 10 percent of studies it was unclear. Finally, with regard to the nature of the treatment interventions, 17 percent were discipline, 17 percent were restorative, 17 percent were counseling, 26 percent were skill building, and 23 percent were multiple types.

The effect size was calculated as the standardized mean difference. The meta-analysis used the inverse of the sample error variance to weight effect sizes.